[perl5.git] / lib / charnames.pm

package charnames;
use strict;
use warnings;
use Carp;
our $VERSION = '1.01';

use bytes ();		# for $bytes::hint_bits
$charnames::hint_bits = 0x20000;

my %alias1 = (
		# Icky 3.2 names with parentheses.
		'LINE FEED'		=> 'LINE FEED (LF)',
		'FORM FEED'		=> 'FORM FEED (FF)',
		'CARRIAGE RETURN'	=> 'CARRIAGE RETURN (CR)',
		'NEXT LINE'		=> 'NEXT LINE (NEL)',
		# Convenience.
		'LF'			=> 'LINE FEED (LF)',
		'FF'			=> 'FORM FEED (FF)',
		'CR'			=> 'CARRIAGE RETURN (LF)',
		'NEL'			=> 'NEXT LINE (NEL)',
		'BOM'			=> 'BYTE ORDER MARK',
	    );

my %alias2 = (
		# Pre-3.2 compatibility (only for the first 256 characters).
		'HORIZONTAL TABULATION'	=> 'CHARACTER TABULATION',
		'VERTICAL TABULATION'	=> 'LINE TABULATION',
		'FILE SEPARATOR'	=> 'INFORMATION SEPARATOR FOUR',
		'GROUP SEPARATOR'	=> 'INFORMATION SEPARATOR THREE',
		'RECORD SEPARATOR'	=> 'INFORMATION SEPARATOR TWO',
		'UNIT SEPARATOR'	=> 'INFORMATION SEPARATOR ONE',
		'PARTIAL LINE DOWN'	=> 'PARTIAL LINE FORWARD',
		'PARTIAL LINE UP'	=> 'PARTIAL LINE BACKWARD',
	    );

my $txt;

# This is not optimized in any way yet
sub charnames
{
  my $name = shift;

  if (exists $alias1{$name}) {
      $name = $alias1{$name};
  }
  if (exists $alias2{$name}) {
      require warnings;
      warnings::warnif('deprecated', qq{Unicode character name "$name" is deprecated, use "$alias2{$name}" instead});
      $name = $alias2{$name};
  }

  my $ord;
  my @off;
  my $fname;

  if ($name eq "BYTE ORDER MARK") {
      $fname = $name;
      $ord = 0xFFFE;
  } else {
      ## Suck in the code/name list as a big string.
      ## Lines look like:
      ##     "0052\t\tLATIN CAPITAL LETTER R\n"
      $txt = do "unicore/Name.pl" unless $txt;

      ## @off will hold the index into the code/name string of the start and
      ## end of the name as we find it.
      
      ## If :full, look for the the name exactly
      if ($^H{charnames_full} and $txt =~ /\t\t\Q$name\E$/m) {
	  @off = ($-[0], $+[0]);
      }

      ## If we didn't get above, and :short allowed, look for the short name.
      ## The short name is like "greek:Sigma"
      unless (@off) {
	  if ($^H{charnames_short} and $name =~ /^(.+?):(.+)/s) {
	      my ($script, $cname) = ($1,$2);
	      my $case = ( $cname =~ /[[:upper:]]/ ? "CAPITAL" : "SMALL");
	      if ($txt =~ m/\t\t\U$script\E (?:$case )?LETTER \U\Q$cname\E$/m) {
		  @off = ($-[0], $+[0]);
	      }
	  }
      }
      
      ## If we still don't have it, check for the name among the loaded
      ## scripts.
      if (not @off)
      {
	  my $case = ( $name =~ /[[:upper:]]/ ? "CAPITAL" : "SMALL");
	  for my $script ( @{$^H{charnames_scripts}} )
	  {
	      if ($txt =~ m/\t\t$script (?:$case )?LETTER \U\Q$name\E$/m) {
		  @off = ($-[0], $+[0]);
		  last;
	      }
	  }
      }
      
      ## If we don't have it by now, give up.
      unless (@off) {
	  carp "Unknown charname '$name'";
	  return "\x{FFFD}";
      }
      
      ##
      ## Now know where in the string the name starts.
      ## The code, in hex, is befor that.
      ##
      ## The code can be 4-6 characters long, so we've got to sort of
      ## go look for it, just after the newline that comes before $off[0].
      ##
      ## This would be much easier if unicore/Name.pl had info in
      ## a name/code order, instead of code/name order.
      ##
      ## The +1 after the rindex() is to skip past the newline we're finding,
      ## or, if the rindex() fails, to put us to an offset of zero.
      ##
      my $hexstart = rindex($txt, "\n", $off[0]) + 1;

      ## we know where it starts, so turn into number -
      ## the ordinal for the char.
      $ord = hex substr($txt, $hexstart, $off[0] - $hexstart);
  }

  if ($^H & $bytes::hint_bits) {	# "use bytes" in effect?
    use bytes;
    return chr $ord if $ord <= 255;
    my $hex = sprintf "%04x", $ord;
    if (not defined $fname) {
	$fname = substr $txt, $off[0] + 2, $off[1] - $off[0] - 2;
    }
    croak "Character 0x$hex with name '$fname' is above 0xFF";
  }

  no warnings 'utf8'; # allow even illegal characters
  return pack "U", $ord;
}

sub import
{
  shift; ## ignore class name

  if (not @_)
  {
      carp("`use charnames' needs explicit imports list");
  }
  $^H |= $charnames::hint_bits;
  $^H{charnames} = \&charnames ;

  ##
  ## fill %h keys with our @_ args.
  ##
  my %h;
  @h{@_} = (1) x @_;

  $^H{charnames_full} = delete $h{':full'};
  $^H{charnames_short} = delete $h{':short'};
  $^H{charnames_scripts} = [map uc, keys %h];

  ##
  ## If utf8? warnings are enabled, and some scripts were given,
  ## see if at least we can find one letter of each script.
  ##
  if (warnings::enabled('utf8') && @{$^H{charnames_scripts}})
  {
      $txt = do "unicore/Name.pl" unless $txt;

      for my $script (@{$^H{charnames_scripts}})
      {
          if (not $txt =~ m/\t\t$script (?:CAPITAL |SMALL )?LETTER /) {
              warnings::warn('utf8',  "No such script: '$script'");
          }
      }
  }
}

require Unicode::UCD; # for Unicode::UCD::_getcode()

my %viacode;

sub viacode
{
    if (@_ != 1) {
        carp "charnames::viacode() expects one numeric argument";
        return ()
    }

    my $arg = shift;
    my $code = Unicode::UCD::_getcode($arg);

    my $hex;

    if (defined $code) {
        $hex = sprintf "%04X", $arg;
    } else {
        carp("unexpected arg \"$arg\" to charnames::viacode()");
        return;
    }

    if ($code > 0x10FFFF) {
	carp "Unicode characters only allocated up to 0x10FFFF (you asked for $hex)";
	return "\x{FFFD}";
    }

    return $viacode{$hex} if exists $viacode{$hex};

    $txt = do "unicore/Name.pl" unless $txt;

    if ($txt =~ m/^$hex\t\t(.+)/m) {
        return $viacode{$hex} = $1;
    } else {
        return;
    }
}

my %vianame;

sub vianame
{
    if (@_ != 1) {
        carp "charnames::vianame() expects one name argument";
        return ()
    }

    my $arg = shift;

    return $vianame{$arg} if exists $vianame{$arg};

    $txt = do "unicore/Name.pl" unless $txt;

    if ($txt =~ m/^([0-9A-F]+)\t\t($arg)/m) {
        return $vianame{$arg} = hex $1;
    } else {
        return;
    }
}


1;
__END__

=head1 NAME

charnames - define character names for C<\N{named}> string literal escapes.

=head1 SYNOPSIS

  use charnames ':full';
  print "\N{GREEK SMALL LETTER SIGMA} is called sigma.\n";

  use charnames ':short';
  print "\N{greek:Sigma} is an upper-case sigma.\n";

  use charnames qw(cyrillic greek);
  print "\N{sigma} is Greek sigma, and \N{be} is Cyrillic b.\n";

  print charname::viacode(0x1234); # prints "ETHIOPIC SYLLABLE SEE"
  printf "%04X", charname::vianame("GOTHIC LETTER AHSA"); # prints "10330"

=head1 DESCRIPTION

Pragma C<use charnames> supports arguments C<:full>, C<:short> and
script names.  If C<:full> is present, for expansion of
C<\N{CHARNAME}}> string C<CHARNAME> is first looked in the list of
standard Unicode names of chars.  If C<:short> is present, and
C<CHARNAME> has the form C<SCRIPT:CNAME>, then C<CNAME> is looked up
as a letter in script C<SCRIPT>.  If pragma C<use charnames> is used
with script name arguments, then for C<\N{CHARNAME}}> the name
C<CHARNAME> is looked up as a letter in the given scripts (in the
specified order).

For lookup of C<CHARNAME> inside a given script C<SCRIPTNAME>
this pragma looks for the names

  SCRIPTNAME CAPITAL LETTER CHARNAME
  SCRIPTNAME SMALL LETTER CHARNAME
  SCRIPTNAME LETTER CHARNAME

in the table of standard Unicode names.  If C<CHARNAME> is lowercase,
then the C<CAPITAL> variant is ignored, otherwise the C<SMALL> variant
is ignored.

Note that C<\N{...}> is compile-time, it's a special form of string
constant used inside double-quoted strings: in other words, you cannot
use variables inside the C<\N{...}>.  If you want similar run-time
functionality, use charnames::vianame().

For the C0 and C1 control characters (U+0000..U+001F, U+0080..U+009F)
as of Unicode 3.1, there are no official Unicode names but you can
use instead the ISO 6429 names (LINE FEED, ESCAPE, and so forth).
In Unicode 3.2 some naming changes will happen since ISO 6429 has been
updated.  Also note that the U+UU80, U+0081, U+0084, and U+0099
do not have names even in ISO 6429.

=head1 CUSTOM TRANSLATORS

The mechanism of translation of C<\N{...}> escapes is general and not
hardwired into F<charnames.pm>.  A module can install custom
translations (inside the scope which C<use>s the module) with the
following magic incantation:

    use charnames ();		# for $charnames::hint_bits
    sub import {
	shift;
	$^H |= $charnames::hint_bits;
	$^H{charnames} = \&translator;
    }

Here translator() is a subroutine which takes C<CHARNAME> as an
argument, and returns text to insert into the string instead of the
C<\N{CHARNAME}> escape.  Since the text to insert should be different
in C<bytes> mode and out of it, the function should check the current
state of C<bytes>-flag as in:

    use bytes ();			# for $bytes::hint_bits
    sub translator {
	if ($^H & $bytes::hint_bits) {
	    return bytes_translator(@_);
	}
	else {
	    return utf8_translator(@_);
	}
    }

=head1 charnames::viacode(code)

Returns the full name of the character indicated by the numeric code.
The example

    print charnames::viacode(0x2722);

prints "FOUR TEARDROP-SPOKED ASTERISK".

Returns undef if no name is known for the code.

This works only for the standard names, and does not yet aply 
to custom translators.

=head1 charnames::vianame(code)

Returns the code point indicated by the name.
The example

    printf "%04X", charnames::vianame("FOUR TEARDROP-SPOKED ASTERISK");

prints "2722".

Returns undef if no name is known for the name.

This works only for the standard names, and does not yet aply 
to custom translators.

=head1 ALIASES

A few aliases have been defined for convenience: instead of having
to use the official names

    LINE FEED (LF)
    FORM FEED (FF)
    CARRIAGE RETURN (CR)
    NEXT LINE (NEL)

(yes, with parentheses) one can use

    LINE FEED
    FORM FEED
    CARRIAGE RETURN
    NEXT LINE
    LF
    FF
    CR
    NEL

One can also use

    BYTE ORDER MARK
    BOM

though that is of course not a legal character as such.

For backward compatibility one can use the old names for
certain C0 and C1 controls

    old                         new

    HORIZONTAL TABULATION       CHARACTER TABULATION
    VERTICAL TABULATION         LINE TABULATION
    FILE SEPARATOR              INFORMATION SEPARATOR FOUR
    GROUP SEPARATOR             INFORMATION SEPARATOR THREE
    RECORD SEPARATOR            INFORMATION SEPARATOR TWO
    UNIT SEPARATOR              INFORMATION SEPARATOR ONE
    PARTIAL LINE DOWN           PARTIAL LINE FORWARD
    PARTIAL LINE UP             PARTIAL LINE BACKWARD

but the old names in addition to giving the character
will also give a warning about being deprecated.

=head1 ILLEGAL CHARACTERS

If you ask for a character that is illegal (like the byte order mark
U+FFFE, or the U+FFFF) does not exist, a warning is given and the
special Unicode I<replacement character> "\x{FFFD}" is returned.

=head1 BUGS

Since evaluation of the translation function happens in a middle of
compilation (of a string literal), the translation function should not
do any C<eval>s or C<require>s.  This restriction should be lifted in
a future version of Perl.

=cut
Commit	Line	Data
423cee85	1	package charnames;
b177ca84 JF	2	use strict;
	3	use warnings;
	4	use Carp;
	5	our $VERSION = '1.01';
b75c8c73	6
d5448623 GS	7	use bytes (); # for $bytes::hint_bits
d5448623 GS	8	$charnames::hint_bits = 0x20000;
423cee85	9
52ea3e69 JH	10	my %alias1 = (
	11	# Icky 3.2 names with parentheses.
	12	'LINE FEED' => 'LINE FEED (LF)',
	13	'FORM FEED' => 'FORM FEED (FF)',
	14	'CARRIAGE RETURN' => 'CARRIAGE RETURN (CR)',
	15	'NEXT LINE' => 'NEXT LINE (NEL)',
	16	# Convenience.
	17	'LF' => 'LINE FEED (LF)',
	18	'FF' => 'FORM FEED (FF)',
	19	'CR' => 'CARRIAGE RETURN (LF)',
51e9e896	20	'NEL' => 'NEXT LINE (NEL)',
52ea3e69 JH	21	'BOM' => 'BYTE ORDER MARK',
	22	);
	23
	24	my %alias2 = (
	25	# Pre-3.2 compatibility (only for the first 256 characters).
	26	'HORIZONTAL TABULATION' => 'CHARACTER TABULATION',
	27	'VERTICAL TABULATION' => 'LINE TABULATION',
	28	'FILE SEPARATOR' => 'INFORMATION SEPARATOR FOUR',
	29	'GROUP SEPARATOR' => 'INFORMATION SEPARATOR THREE',
	30	'RECORD SEPARATOR' => 'INFORMATION SEPARATOR TWO',
	31	'UNIT SEPARATOR' => 'INFORMATION SEPARATOR ONE',
	32	'PARTIAL LINE DOWN' => 'PARTIAL LINE FORWARD',
	33	'PARTIAL LINE UP' => 'PARTIAL LINE BACKWARD',
	34	);
	35
423cee85 JH	36	my $txt;
	37
	38	# This is not optimized in any way yet
b177ca84 JF	39	sub charnames
	40	{
	41	my $name = shift;
	42
52ea3e69 JH	43	if (exists $alias1{$name}) {
	44	$name = $alias1{$name};
	45	}
	46	if (exists $alias2{$name}) {
	47	require warnings;
	48	warnings::warnif('deprecated', qq{Unicode character name "$name" is deprecated, use "$alias2{$name}" instead});
	49	$name = $alias2{$name};
	50	}
b177ca84	51
52ea3e69	52	my $ord;
423cee85	53	my @off;
52ea3e69 JH	54	my $fname;
	55
	56	if ($name eq "BYTE ORDER MARK") {
	57	$fname = $name;
	58	$ord = 0xFFFE;
	59	} else {
	60	## Suck in the code/name list as a big string.
	61	## Lines look like:
	62	## "0052\t\tLATIN CAPITAL LETTER R\n"
	63	$txt = do "unicore/Name.pl" unless $txt;
b177ca84	64
52ea3e69 JH	65	## @off will hold the index into the code/name string of the start and
	66	## end of the name as we find it.
	67
	68	## If :full, look for the the name exactly
	69	if ($^H{charnames_full} and $txt =~ /\t\t\Q$name\E$/m) {
	70	@off = ($-[0], $+[0]);
423cee85	71	}
b177ca84	72
52ea3e69 JH	73	## If we didn't get above, and :short allowed, look for the short name.
	74	## The short name is like "greek:Sigma"
	75	unless (@off) {
	76	if ($^H{charnames_short} and $name =~ /^(.+?):(.+)/s) {
	77	my ($script, $cname) = ($1,$2);
	78	my $case = ( $cname =~ /[[:upper:]]/ ? "CAPITAL" : "SMALL");
	79	if ($txt =~ m/\t\t\U$script\E (?:$case )?LETTER \U\Q$cname\E$/m) {
	80	@off = ($-[0], $+[0]);
	81	}
	82	}
	83	}
	84
	85	## If we still don't have it, check for the name among the loaded
	86	## scripts.
	87	if (not @off)
b177ca84	88	{
52ea3e69 JH	89	my $case = ( $name =~ /[[:upper:]]/ ? "CAPITAL" : "SMALL");
	90	for my $script ( @{$^H{charnames_scripts}} )
	91	{
	92	if ($txt =~ m/\t\t$script (?:$case )?LETTER \U\Q$name\E$/m) {
	93	@off = ($-[0], $+[0]);
	94	last;
	95	}
	96	}
b177ca84	97	}
52ea3e69 JH	98
	99	## If we don't have it by now, give up.
	100	unless (@off) {
	101	carp "Unknown charname '$name'";
	102	return "\x{FFFD}";
	103	}
	104
	105	##
	106	## Now know where in the string the name starts.
	107	## The code, in hex, is befor that.
	108	##
	109	## The code can be 4-6 characters long, so we've got to sort of
	110	## go look for it, just after the newline that comes before $off[0].
	111	##
	112	## This would be much easier if unicore/Name.pl had info in
	113	## a name/code order, instead of code/name order.
	114	##
	115	## The +1 after the rindex() is to skip past the newline we're finding,
	116	## or, if the rindex() fails, to put us to an offset of zero.
	117	##
	118	my $hexstart = rindex($txt, "\n", $off[0]) + 1;
	119
	120	## we know where it starts, so turn into number -
	121	## the ordinal for the char.
	122	$ord = hex substr($txt, $hexstart, $off[0] - $hexstart);
423cee85	123	}
b177ca84	124
d5448623	125	if ($^H & $bytes::hint_bits) { # "use bytes" in effect?
8058d7ab	126	use bytes;
d41ff1b8	127	return chr $ord if $ord <= 255;
f0175764	128	my $hex = sprintf "%04x", $ord;
52ea3e69 JH	129	if (not defined $fname) {
	130	$fname = substr $txt, $off[0] + 2, $off[1] - $off[0] - 2;
	131	}
f0175764	132	croak "Character 0x$hex with name '$fname' is above 0xFF";
423cee85	133	}
f0175764	134
52ea3e69	135	no warnings 'utf8'; # allow even illegal characters
bfa383d6	136	return pack "U", $ord;
423cee85 JH	137	}
423cee85 JH	138
b177ca84 JF	139	sub import
	140	{
	141	shift; ## ignore class name
	142
	143	if (not @_)
	144	{
	145	carp("`use charnames' needs explicit imports list");
	146	}
d5448623	147	$^H \|= $charnames::hint_bits;
423cee85	148	$^H{charnames} = \&charnames ;
b177ca84 JF	149
	150	##
	151	## fill %h keys with our @_ args.
	152	##
423cee85 JH	153	my %h;
423cee85 JH	154	@h{@_} = (1) x @_;
b177ca84	155
423cee85 JH	156	$^H{charnames_full} = delete $h{':full'};
	157	$^H{charnames_short} = delete $h{':short'};
	158	$^H{charnames_scripts} = [map uc, keys %h];
b177ca84 JF	159
	160	##
	161	## If utf8? warnings are enabled, and some scripts were given,
	162	## see if at least we can find one letter of each script.
	163	##
	164	if (warnings::enabled('utf8') && @{$^H{charnames_scripts}})
	165	{
	166	$txt = do "unicore/Name.pl" unless $txt;
	167
	168	for my $script (@{$^H{charnames_scripts}})
	169	{
	170	if (not $txt =~ m/\t\t$script (?:CAPITAL \|SMALL )?LETTER /) {
	171	warnings::warn('utf8', "No such script: '$script'");
	172	}
	173	}
bd62941a	174	}
423cee85 JH	175	}
423cee85 JH	176
f0175764 JH	177	require Unicode::UCD; # for Unicode::UCD::_getcode()
f0175764 JH	178
4e2cda5d JH	179	my %viacode;
4e2cda5d JH	180
b177ca84 JF	181	sub viacode
	182	{
	183	if (@_ != 1) {
daf0d493	184	carp "charnames::viacode() expects one numeric argument";
b177ca84 JF	185	return ()
b177ca84 JF	186	}
f0175764	187
b177ca84	188	my $arg = shift;
f0175764	189	my $code = Unicode::UCD::_getcode($arg);
b177ca84 JF	190
b177ca84 JF	191	my $hex;
f0175764 JH	192
f0175764 JH	193	if (defined $code) {
b177ca84 JF	194	$hex = sprintf "%04X", $arg;
	195	} else {
	196	carp("unexpected arg \"$arg\" to charnames::viacode()");
daf0d493	197	return;
b177ca84 JF	198	}
b177ca84 JF	199
f0175764 JH	200	if ($code > 0x10FFFF) {
	201	carp "Unicode characters only allocated up to 0x10FFFF (you asked for $hex)";
	202	return "\x{FFFD}";
	203	}
	204
4e2cda5d JH	205	return $viacode{$hex} if exists $viacode{$hex};
4e2cda5d JH	206
b177ca84 JF	207	$txt = do "unicore/Name.pl" unless $txt;
	208
	209	if ($txt =~ m/^$hex\t\t(.+)/m) {
4e2cda5d	210	return $viacode{$hex} = $1;
b177ca84	211	} else {
daf0d493 JH	212	return;
	213	}
	214	}
	215
4e2cda5d JH	216	my %vianame;
4e2cda5d JH	217
daf0d493 JH	218	sub vianame
	219	{
	220	if (@_ != 1) {
	221	carp "charnames::vianame() expects one name argument";
	222	return ()
	223	}
	224
	225	my $arg = shift;
	226
4e2cda5d JH	227	return $vianame{$arg} if exists $vianame{$arg};
4e2cda5d JH	228
daf0d493 JH	229	$txt = do "unicore/Name.pl" unless $txt;
	230
	231	if ($txt =~ m/^([0-9A-F]+)\t\t($arg)/m) {
4e2cda5d	232	return $vianame{$arg} = hex $1;
daf0d493 JH	233	} else {
daf0d493 JH	234	return;
b177ca84 JF	235	}
	236	}
	237
423cee85 JH	238
	239	1;
	240	__END__
	241
	242	=head1 NAME
	243
b177ca84	244	charnames - define character names for C<\N{named}> string literal escapes.
423cee85 JH	245
	246	=head1 SYNOPSIS
	247
	248	use charnames ':full';
4a2d328f	249	print "\N{GREEK SMALL LETTER SIGMA} is called sigma.\n";
423cee85 JH	250
423cee85 JH	251	use charnames ':short';
4a2d328f	252	print "\N{greek:Sigma} is an upper-case sigma.\n";
423cee85 JH	253
423cee85 JH	254	use charnames qw(cyrillic greek);
4a2d328f	255	print "\N{sigma} is Greek sigma, and \N{be} is Cyrillic b.\n";
423cee85	256
b177ca84	257	print charname::viacode(0x1234); # prints "ETHIOPIC SYLLABLE SEE"
daf0d493	258	printf "%04X", charname::vianame("GOTHIC LETTER AHSA"); # prints "10330"
b177ca84	259
423cee85 JH	260	=head1 DESCRIPTION
	261
	262	Pragma C<use charnames> supports arguments C<:full>, C<:short> and
	263	script names. If C<:full> is present, for expansion of
4a2d328f	264	C<\N{CHARNAME}}> string C<CHARNAME> is first looked in the list of
423cee85 JH	265	standard Unicode names of chars. If C<:short> is present, and
	266	C<CHARNAME> has the form C<SCRIPT:CNAME>, then C<CNAME> is looked up
	267	as a letter in script C<SCRIPT>. If pragma C<use charnames> is used
4a2d328f	268	with script name arguments, then for C<\N{CHARNAME}}> the name
423cee85 JH	269	C<CHARNAME> is looked up as a letter in the given scripts (in the
	270	specified order).
	271
	272	For lookup of C<CHARNAME> inside a given script C<SCRIPTNAME>
d5448623	273	this pragma looks for the names
423cee85 JH	274
	275	SCRIPTNAME CAPITAL LETTER CHARNAME
	276	SCRIPTNAME SMALL LETTER CHARNAME
	277	SCRIPTNAME LETTER CHARNAME
	278
	279	in the table of standard Unicode names. If C<CHARNAME> is lowercase,
daf0d493 JH	280	then the C<CAPITAL> variant is ignored, otherwise the C<SMALL> variant
	281	is ignored.
	282
	283	Note that C<\N{...}> is compile-time, it's a special form of string
	284	constant used inside double-quoted strings: in other words, you cannot
4e2cda5d	285	use variables inside the C<\N{...}>. If you want similar run-time
daf0d493	286	functionality, use charnames::vianame().
423cee85	287
301a3cda JH	288	For the C0 and C1 control characters (U+0000..U+001F, U+0080..U+009F)
	289	as of Unicode 3.1, there are no official Unicode names but you can
	290	use instead the ISO 6429 names (LINE FEED, ESCAPE, and so forth).
	291	In Unicode 3.2 some naming changes will happen since ISO 6429 has been
	292	updated. Also note that the U+UU80, U+0081, U+0084, and U+0099
	293	do not have names even in ISO 6429.
	294
423cee85 JH	295	=head1 CUSTOM TRANSLATORS
423cee85 JH	296
d5448623	297	The mechanism of translation of C<\N{...}> escapes is general and not
423cee85	298	hardwired into F<charnames.pm>. A module can install custom
d5448623	299	translations (inside the scope which C<use>s the module) with the
423cee85 JH	300	following magic incantation:
423cee85 JH	301
d5448623 GS	302	use charnames (); # for $charnames::hint_bits
	303	sub import {
	304	shift;
	305	$^H \|= $charnames::hint_bits;
	306	$^H{charnames} = \&translator;
	307	}
423cee85 JH	308
	309	Here translator() is a subroutine which takes C<CHARNAME> as an
	310	argument, and returns text to insert into the string instead of the
4a2d328f	311	C<\N{CHARNAME}> escape. Since the text to insert should be different
d5448623 GS	312	in C<bytes> mode and out of it, the function should check the current
	313	state of C<bytes>-flag as in:
	314
	315	use bytes (); # for $bytes::hint_bits
	316	sub translator {
	317	if ($^H & $bytes::hint_bits) {
	318	return bytes_translator(@_);
	319	}
	320	else {
	321	return utf8_translator(@_);
	322	}
423cee85	323	}
423cee85	324
b177ca84 JF	325	=head1 charnames::viacode(code)
	326
	327	Returns the full name of the character indicated by the numeric code.
	328	The example
	329
	330	print charnames::viacode(0x2722);
	331
	332	prints "FOUR TEARDROP-SPOKED ASTERISK".
	333
daf0d493 JH	334	Returns undef if no name is known for the code.
	335
	336	This works only for the standard names, and does not yet aply
	337	to custom translators.
	338
	339	=head1 charnames::vianame(code)
	340
	341	Returns the code point indicated by the name.
	342	The example
	343
	344	printf "%04X", charnames::vianame("FOUR TEARDROP-SPOKED ASTERISK");
	345
	346	prints "2722".
	347
	348	Returns undef if no name is known for the name.
b177ca84 JF	349
	350	This works only for the standard names, and does not yet aply
	351	to custom translators.
	352
52ea3e69 JH	353	=head1 ALIASES
	354
	355	A few aliases have been defined for convenience: instead of having
	356	to use the official names
	357
	358	LINE FEED (LF)
	359	FORM FEED (FF)
	360	CARRIAGE RETURN (CR)
	361	NEXT LINE (NEL)
	362
	363	(yes, with parentheses) one can use
	364
	365	LINE FEED
	366	FORM FEED
	367	CARRIAGE RETURN
	368	NEXT LINE
	369	LF
	370	FF
	371	CR
	372	NEL
	373
	374	One can also use
	375
	376	BYTE ORDER MARK
	377	BOM
	378
	379	though that is of course not a legal character as such.
	380
	381	For backward compatibility one can use the old names for
	382	certain C0 and C1 controls
	383
	384	old new
	385
	386	HORIZONTAL TABULATION CHARACTER TABULATION
	387	VERTICAL TABULATION LINE TABULATION
	388	FILE SEPARATOR INFORMATION SEPARATOR FOUR
	389	GROUP SEPARATOR INFORMATION SEPARATOR THREE
	390	RECORD SEPARATOR INFORMATION SEPARATOR TWO
	391	UNIT SEPARATOR INFORMATION SEPARATOR ONE
	392	PARTIAL LINE DOWN PARTIAL LINE FORWARD
	393	PARTIAL LINE UP PARTIAL LINE BACKWARD
	394
	395	but the old names in addition to giving the character
	396	will also give a warning about being deprecated.
	397
f0175764 JH	398	=head1 ILLEGAL CHARACTERS
f0175764 JH	399
52ea3e69 JH	400	If you ask for a character that is illegal (like the byte order mark
	401	U+FFFE, or the U+FFFF) does not exist, a warning is given and the
	402	special Unicode I<replacement character> "\x{FFFD}" is returned.
f0175764	403
423cee85 JH	404	=head1 BUGS
	405
	406	Since evaluation of the translation function happens in a middle of
	407	compilation (of a string literal), the translation function should not
	408	do any C<eval>s or C<require>s. This restriction should be lifted in
	409	a future version of Perl.
	410
	411	=cut